Adequately trained EEG technicians are essential for proper electrode placement and appropriate recording.[8][9][10] They should be familiar with artifactual background activity, prompt correction of misplaced electrodes, or adjust settings. Bias and artifactual background should be minimized. An interdisciplinary team that includes neonatal critical care nurses, primary care providers, consultants, technicians, and respiratory therapists must coordinate care and obtain the most accurate and precise data from the continuous video EEG.

During a neonatal EEG recording, the following steps are essential:

• Minimize movement or stimulation of the patient
• Appropriate equipment, electrodes, and tools are available
• Documentation of topography (head positioning), gestational age, conceptional age, state of the infant, among other relevant neonatal data.
• Minimize artifacts
• Adequate pasting of the skin to avoid breakdown; keep the impedance < 10 kOhms

The neonatal EEG total number of electrodes is reduced, given the small head circumference.[2] Usually, a longitudinal or average montage is used, with a standard 10-20 electrode system. Electrodes are connected in frontotemporal, frontocentral, temporal-occipital, and central-occipital measurements. The interelectrode distance is 40% from the diameter of the head for neonatal recordings. Physiological leads include an EMG, EKG, and respiratory monitors; they can distinguish artifactual from epileptiform activity.[7]

The eye electrodes are placed 0.5 cm from the outer canthus. Calibration of montage is necessary to eliminate artifact with a recording speed of 15 mm/s (half from adults), a time constant of 0.3 seconds, 70 Hz low pass filter, and sensitivity of 7 µV/mm. EKG, EMG, and eye electrodes are set at sensitivities of 50 µV/mm, 50 µV/mm, and 7 µV/mm, respectively. These settings may be adjusted as needed and are meant to be guidance. Abnormal sharp spike and slow waves should disrupt the background and be sharply contoured. A recording of at least 2 to 3 hours is recommended to capture waves in the wake and the sleep stages.[2][7]

The neonatal EEG voltage can be decreased by cooling therapy (head or body cooling) and some medications, including benzodiazepines, barbiturates, morphine, and antiepileptic drugs.[7] Eye movements and respirations help to differentiate the sleep stages. Artifacts in the EEG can be produced by EKG, pulse, nursing care, parental care, respirations, ventilator, incubator, twitches, loud noise, and feeding.

Complications are common during extensive monitoring and include the following:

• Skin breakdown 
• Skin infections

The use of moisturizers, gel, and topical antibiotics can preserve skin and avoid infections.

Introduction and Obstetric Definitions

Neonatal seizures or convulsions are epileptic events occurring from birth to the end of the neonatal period, which is about 28 days. The neonatal period is the most vulnerable period for the development of seizures, with an incidence of 80% in the first 24 to 48 hours of life. The corrected age should be calculated to interpret a neonatal EEG. Benign variants will be reported according to the corrected age and not the chronological age of the neonate. The first trimester runs from conception to about 12 weeks, the second trimester 13-26 weeks, and the third trimester 27 to 42 weeks.

Visual analysis inspects the background rhythm for continuity/discontinuity, symmetry, synchrony, amplitude, reactivity, and morphology.[7] A continuity pattern is constant and with a similar amplitude, while the discontinuity pattern has high amplitude burst mixed with low amplitude interburst. Background varies with age becoming more continuous, synchronous, and reactive to stimuli with advancing gestational age. Active sleep corresponds to REM sleep, while quiet sleep corresponds to non-REM sleep. Symmetry compares homologous brain regions in both hemispheres for amplitude changes. Synchrony refers to bursts occurring in the right and left hemisphere separated no more than 1.5 seconds. During 31 and 36 weeks, the bursts are 70% synchronous compared to the periods before 31 weeks or after 36 weeks when it is 100% synchronous. The amplitude decreases from 24 weeks until term.

Classification

A neonatal epilepsy disorder can point towards a more ominous sign of sustained and permanent brain damage. Early detection, management, and aggressive treatment of seizures and status epilepticus can prevent long term neurological sequela. About 3/1000 neonates will have a seizure disorder in the first two days of life in term babies and up to 132/1000 in pre-term infants. Neonatal seizures can be classified as subtle, tonic, clonic, generalized tonic-clonic, myoclonic, with behavioral features that can be paroxysmal, non-paroxysmal, repetitive, and stereotypical. 

• Subtle: (approximately 50%) Difficult to detect clinically. It can consist of aleatory roving eye movements, including open fluttering and rolling up. Oral smacking or automatisms, pedal movements, staring/apneic events, repetitive sucking, chewing sounds, and tongue protrusion.
• Clonic: (approximately 25%)  Involuntary, rhythmic muscle contractions that are fast frequency.
• Myoclonic: (approximately 20%) Repetitive brief shock-like jerks of the limbs.
• Tonic: (approximately 5%) Repetitive and sustained tonic/stiffening flexion or extension of the limbs, with decerebrate and decorticate like posturing.
• Generalized tonic-clonic: Repetitive alternating tonic-clonic movements that represent a primary or secondarily generalized seizure associated with poor prognosis, but very rare.

EEG Patterns

EEG patterns can vary depending on the gestational and corrected age. Features to identify in the EEG include changes in posterior dominant rhythm, frequency, wave amplitude, synchrony of burst of activity, sleep wave activity, and percentage of non-REM/REM stages.

1. Clinical EEG on neonatal stages of sleep[2][7][11]

• Awake
  • Open eyes characterize it with rapid movements, irregular and active respiration, rapid startles, and gross appendicular movements. Normal behavior includes smiling, grimacing, sucking, vocalization such as crying, and startle.
  • EEG shows fast frequency, continuous medium voltage (70-100 µV) wave morphology, with a narrow bandwidth on amplitude-integrated EEG.
• Quiet Sleep
  • It is characterized by closed eyes, lack of eye movements, atonia, and regular respiration. May have normal rhythmic mouth movements. REM waves may or may not be seen depending on gestational age. EEG shows slow frequency, trace alternant, discontinuous, with medium/high voltage (30 to 200 µV) wave morphology, at times with a wide bandwidth on amplitude-integrated EEG.
• Active Sleep
  • It is characterized by REM, atonia, occasional frowns, vocalizations, or sucking behavior while asleep.
  • EEG may show continuous low voltage (30 to 70 µV) wave morphology, has a narrow bandwidth on amplitude-integrated EEG.
• Transitional Sleep
  • Characterized by alternation between open eyes and closed eyes, slow eye movements, difficult to startle, at times grimacing, intermittent sucking, and increase sleep vocalizations may be present.
  • EEG shows alternating continuous and discontinuous, high voltage (100 to 200 µV) wave morphology, at times with variable bandwidth on amplitude-integrated EEG.

2. EEG patterns per gestational age

• 24-28 weeks
  • Discontinuous background, monophasic or diphasic delta with superimposed theta frequency.
  • Delta activity waves have a high amplitude (up to 300 µV) occurring in multiple cortical areas, sometimes excluding frontal areas.
  • Theta bursts are frequent in bitemporal areas and become predominant at the end of this phase.
  • Stimulation produces no reactivity.
• 28-31 weeks
  • Sleep stages become more defined and easier to differentiate by discontinuous and continuous rapid eye movements.
  • Delta activity is focal, the synchronous frequency with localization to the central and occipital regions, and predominant in an awake state. It has a smaller amplitude (200 µV) with the presence of superimposed alpha or theta rhythms.
  • Synchronized theta bursts occur in temporal-occipital areas in the quiet stage of sleep. Prominent delta brushes (delta waves with superimposed alpha or beta rhythms) are common.
  • Stimulus induced reactivity and attenuation of the background can be elicited.
• 32-34 weeks
  • This age entirely defines sleep structures with a continuous wave morphology in wakefulness and active sleep, and discontinuous in quiet sleep.
  • Increased delta burst duration and frequency, decreased inter-burst intervals (e.g., < 15 seconds at 32 weeks, and < 10 sec at 34 weeks respectively), and reduction in wave amplitude.
  • Delta brush activity is localized to the occipital by 34 weeks.
  • Theta activity may be seen but disappears at the end of this stage.
  • Frontal sharp transients may be present.
• 35-36 weeks
  • Sleep structures are defined.
    • Delta brushes are defined in the awake state. The appearance of discontinuous background activity characterizes quiet sleep, and REM is characterized by the appearance of continuous/low amplitude waveforms with theta wave activity.
    • Background activity is continuous, bisynchronous, variable frequencies in active sleep, and asynchronous in quiet sleep.
  • Continuous high amplitude, slow frequency (~1-3 Hz) delta bursts can be seen.
• 37-44 weeks
  • Sleep stages are defined.
  • Background activity remains unchanged with variable frequencies, high amplitude, and evident changes by sleep stages, including trace alternant in quiet sleep.
  • Localized occipital delta brushes until 40 weeks.
  • Frontal sharp transients and anterior slow dysrhythmia began to disappear as the patient approaches 44 weeks.
• 44-46 weeks
  • Background activity is continuous in all states.
  • Sleep spindles are seen.

An interdisciplinary neonatal team of pediatric neurologists, clinical pediatric neurophysiologists, pediatric critical care specialists, neonatal nurses, and respiratory therapists should work together to achieve an optimal video EEG recording. Documentation of clinical symptoms, current medications, topography, gestational age, corrected age, and state of the infant, among other clinical data, is essential and should be conveyed to the EEG reader for accurate study interpretation. Protocols should be in place to assist and facilitate coordination of care. [Level 1]